Live wire Humanoids · Autonomous vehicles · Drones · Industrial robotics · Embodied AI research

Can Robots Bring Stroke Surgery to Remote Hospitals?

Written by

in

I am practicing brain surgery on a patient lying 6 meters away, out of my sight and reach. I press a foot pedal, activating the controller in my hand, and carefully steer a thin flexible tube called a catheter up through the patient’s blood vessels. I head for what I believe to be the brain, tracking my progress on the live X-ray feed. As I imagine finding a blood clot, David Bell, the surgeon seated to my right, helpfully reminds me that every minute a stroke goes untreated, millions of brain cells suffocate, swell, and die.The patient under the X-ray in Bell’s office is not flesh and blood, but a plastic container holding long branches of silicone tubing. The controller I hold comes from an Xbox. But the surgical robot is real. The Remedy N1 is designed for emergency operations, including thrombectomies, the standard brain operation for acute ischemic strokes caused by blood clots. Ischemia, a restriction of blood flow, is the most common cause of stroke.“We want it to feel a bit like a video game—a bit easy,” Bell says. “But we don’t want to dehumanize it as if there’s just another ‘game over.’ ” Bell, a clinician and former cardiac-surgery trainee from Australia, cofounded Remedy Robotics, based in San Francisco. With a team of engineers, he developed this robotic catheter and its related software to help ischemic stroke patients in remote and rural places, who lose crucial hours while being transferred away from hospitals incapable of providing care. According to 2025 data, at least 5 million people around the world suffer ischemic strokes every year. In the United States, stroke is the fourth leading cause of death. When Remedy Robotics cofounders Jake Sganga [left] and David Bell initially met, they matched each other’s energy right away.Christie Hemm KlokThe screen shows my catheter at a fork in the narrowing road, and I decide to take a sharp right; I think I can make it. This device feels easier to drive than a typical robotic instrument that I controlled on a hospital tour the summer prior. I say to Bell that here my movements are smoother, more precise—exactly the talking point you’d expect to hear from a surgical robotics company. Never mind that, Bell corrects me: “Smoothness and precision are not clinical outcomes. Neither is a surgeon’s ergonomics.” For Bell, the outcome that matters is whether the robot can get stroke treatment to patients who otherwise wouldn’t receive it in time. Today, only about 1 in 10 patients eligible for thrombectomy receive one. That’s where the long-unrealized promise of telesurgery is meeting an urgent problem: providing access to thrombectomies before life-threatening damage takes hold. But can putting a robot in one hospital and a surgeon somewhere else actually close that gap? Remedy has already taken a first step, remotely testing its system on a small group of patients. “The remote stroke need is well known,” says Jake Sganga, who cofounded Remedy with Bell and leads the technology side. “It’s just a hard problem to take on.” Making telesurgery a practical part of stroke care will require more than a capable robot: It will depend on communications infrastructure, hospital resources and workflows, and ultimately, evidence that patients actually benefit.Why Telesurgery Starts With StrokeAs recently as the early ’90s, physicians who treated stroke struggled with “therapeutic nihilism.” A 1991 editorial by two lamenting neurologists expressed hope that a successful drug therapy for stroke would emerge that would “reverse the trend of nihilism.” Five years later, clot-busting drugs revolutionized stroke treatment, followed about a decade later by mechanical thrombectomy, in which a blood clot is physically removed from an artery.In a standard thrombectomy a surgeon snakes a flexible wire through a small opening in the patient’s groin. This guide wire sets the route for a fine hollow tube around it, the catheter, to travel through the blood vessels to the brain, where instruments remove the clot. Suddenly a stroke was no longer a sentence of profound disability or death but a treatable event.Yet decades later, thrombectomies remain inherently blind expeditions. The surgeon stands beside the patient, physically pushing these lines up, inch by inch. The only view comes from a live X-ray, which can’t see soft tissues like blood vessels. The surgeon shoots brief puffs of contrast dye through the catheter to illuminate the path ahead. A nimble operator can avoid scraping or piercing artery walls while bending the guide wire’s rounded tip around corners. But even these left or right turns are not intuitive. The guide wire can twist inside the body, and because the surgeon sees only a flat silhouette, steering it left may actually rotate the guidewire to the right. The N1 robot feeds the catheter along a rail lined with seven modules, using a thin wire that allows precise control through a patient’s vascular system, even from kilometers away. Christie Hemm Klok If all goes as planned, the doctor reaches the brain and pushes instruments through the catheter to dislodge the clot, then carefully pulls everything back out. Bell remembers how thrombectomies felt when he performed them. “I’m one and a half meters away from my catheter tip. I’m not in control of it. I can’t see where the blood vessels are. I thought, what is the ideal way to do these procedures? It’s certainly not what I’m doing now.”Sganga and Bell want their robots to do two things: see what surgeons can’t see and be where surgeons can’t be. When I visited Remedy’s office in January, the team showed me their robotic arm and instruments that snap onto it. Motorized cartridges live on the robotic arm, one containing several actuators that use small gears to grip the threadlike instruments to move them in and out of a body. The instruments include a stent, an expandable wire-mesh tube used to pierce, grab, and yank a clot back out. It’s all typical stroke operation equipment—except for the catheter.Remedy’s catheter is steerable, so it doesn’t need a guide wire to pave its path. Outside the body, a robotic arm advances the catheter while tiny wires tug its leading tip in any direction, as if an inquisitive worm is navigating the narrow branches of a bloodstream. As I maneuver the catheter with the Xbox controller, Bell and Sganga tell me they made wayfinding a priority. Sganga toggles on a CT scan to overlay on the X-ray feed. Suddenly I see the bright three-dimensional routes of blood vessels along a path that had been invisible on X-ray.I noticed a pattern of small markings on the catheter. Sganga tells me they’re not for the surgeon, but for a computer-vision algorithm that tracks the catheter’s rotation inside the body. If an operator wants to steer right, Remedy’s software automatically translates that intention into the correct action. Instead of squinting at a 2D silhouette on an X-ray and mentally compensating for every twist in the catheter, the surgeon simply toggles to the right on the X-ray image, the Remedy system adjusts, and the catheter moves in the correct direction. This system grew out of ideas that Bell and Sganga were pursuing when they met at Stanford in 2019. Already a physician, Bell was completing an MBA and enrolled in a robotics class in hopes of meeting an engineer who might share his interest in creating a medical robot. The professor connected him to Sganga, a bioengineering doctoral student who was developing pathfinding algorithms for an FDA-approved robotic bronchoscope to navigate the lungs.Outside the body, a robotic arm advances the catheter while tiny wires tug its leading tip in any direction, as if an inquisitive worm is navigating the narrow branches of a bloodstream.Sganga’s engineering work had exposed him to surgical robotics companies that failed because they hadn’t chosen a valuable enough problem to solve. He sought a medical problem where a robot could prove its worth well before human trials, and stroke stood out. “It’s an already verified clinical need,” Sganga says, “Time.” When he and Bell met for coffee, they matched each other’s energy right away: “It was like ‘Hey how you doing?’ and then we were looking at medical images of blood vessels together off of his phone,” Sganga says.About a year after they both left Stanford, Bell and Sganga rented a two-car garage in Southern California and X-ray-proofed it with lead lining. “We called it the bunker because it had exposed lead on all sides, including the ceiling,” Sganga says. “It was just big enough to fit a bed and this rickety old C-arm [X-ray].” There they test-drove a prototype of their steerable catheter through silicone arteries—first controlling the robot on-site from another room, and then from a laptop hundreds of kilometers away. “We were trying to get an initial sign of life,” Sganga says. “And it was that moment that we realized that this could work.” Telesurgery’s Second ActDreams of remote surgery predate modern stroke care. In the 1980s and 1990s, the companies Computer Motion and Intuitive developed bedside robots to assist surgeons. Then, in 2001, French surgeon Jacques Marescaux and Canadian surgeon Michel Gagner removed a patient’s gallbladder from more than 6,000 km away. In New York City, Marescaux used a computer connected via high-speed fiber optics to control a Computer Motion robot in Strasbourg, France. The milestone, known as the Lindbergh Operation, was completed safely with no complications.But telesurgery did not take off after Lindbergh. Instead, companies invested in robots operated by surgeons in the same room. Computer Motion and Intuitive merged in 2003, more companies sprang up, and robotic surgery steadily entered the mainstream. Today thousands of surgical robots operate in hospitals around the country, assisting in roughly 22 percent of surgeries. In 2023 alone, more than 2 million surgeries were performed with da Vinci, Intuitive’s flagship platform. Remedy mechanical engineer Chris Hapstack [left] and software engineer Campell Crowley adjust the N1 robot. Christie Hemm KlokIn the summer of 2025, telesurgery appeared to regain momentum when surgeons near Orlando, Fla., removed the prostate from a man 11,000 km away in Luanda, Angola. Around the same time, Intuitive demonstrated what it called “telecollaboration,” in which two surgeons on different continents operated together on a tissue model. Meanwhile, India-based SS Innovations has been testing cardiac telesurgery; in Panama, the Texas-based medical robotics company XCath has tested a steerable guide wire for thrombectomies. These companies are building on advances in computer vision and machine learning, yet remote robotics is hardly a sure bet.“The first procedures are going to be those procedures where time is critical,” says Brad Nelson, who develops surgical robots at ETH Zurich and is not affiliated with Remedy. Last year, Nelson coauthored a report in Science Robotics arguing that telesurgery’s future success hinges on finding emergency operations where patients can’t always reach specialists in time, including stroke, heart attack, and fetal operations.Surgeries that occur within blood vessels seem an attractive proving ground for teleoperation in emergencies. “If you have a hiccup, the nice thing about vascular procedures is nothing bad can happen other than the thing stops moving,” Nelson says. There’s neither a large open cavity to tend to, nor the urgent choreography one might need in organ transplants. If the surgeon loses connection to the robot while the catheter is inside a blood vessel, the operation can pause in place without damage.Nelson’s lab is also working on a stroke solution. He has partnered with the Mayo Clinic in Arizona to build microscopic robots—the size of a coffee grain—steered with magnets to dissolve a clot. When Surgery Depends on the NetworkAfter preliminary success in the garage, Sganga and Bell turned their focus to the environment where the Remedy N1 would live. Doctors and nurses advised them to improve the live audio-video feed, as any surgery is a coordinated team effort requiring constant communication. They were told to improve setup time, maintenance, cleaning, and began a series of conversations about secure communication and cybersecurity. In 2021, CEO David Bell naps next to Remedy’s first X-ray machine in a lead-lined garage in Venice, Calif. Remedy RoboticsAbsent a cyberattack, the elephant in the room for any telesurgery hopeful is latency. Experts speak about latency, or lag, in terms of how many thousandths of a second it takes for a small packet of data to reach a destination then ping back to its original source. In a major city with strong 5G or fiber-optic internet, latency might measure in the tens of milliseconds. In rural areas, even in the United States, lag can drag above 100.Noticeable lag may not be a dealbreaker. When researchers in Japan analyzed five years of surgical simulations and demonstrations, they reported surgeons adapting to the connection speed. They paused. They moved more slowly. Remedy has noticed similar behavior in test runs. “We’ve actually found that operators are quite resilient to latency,” Sganga says.Still, Remedy targets a round-trip latency of no more than 150 milliseconds, below the average human reaction time of 200 milliseconds. Whenever latency sneaks above that threshold, the system pauses all surgical action. Bell and Sganga say they’ve tailored their system to work with the slower internet speeds common in rural areas that would benefit from better access to stroke care, claiming they recently operated their system from across the Pacific using a mobile hotspot. But sorting out the engineering details raises more questions about how telesurgery might change surgical practice. Bell told me he worries about what a telesurgical future could bring. What if surgeons’ skills erode? What if trainees become dependent on specific platforms? What if two patients need care from the same remote surgeon at the same time? “Is the person suffering a stroke in America necessarily more important than the person suffering a stroke in India? No,” Bell says. “So how do you figure that out? That’s something I think about a lot.”Where Telesurgery Could Work BestDuring a stroke, a blockage in the brain triggers symptoms within seconds or minutes. The following hour is the so-called golden period for saving the brain before the risk of permanent damage grows. “Every minute wasted or delayed, 2 million brain cells are going to die,” says neurologist Archit Bhatt of Providence Medical Group in Portland, Oregon. “You really need to act fast.” Yet decades after the advent of thrombectomies, one in five Americans lives more than an hour from a hospital that offers the procedure.Medium and small hospitals often lack the 24-hour surgical team, experience, or equipment for thrombectomies, so they must transfer stroke patients to large hospitals. As a result, standard stroke care is concentrated in urban areas. For example, Alaska’s only two thrombectomy centers serving the 739,000 people who live or work in the state are both in Anchorage, leaving more than half of residents out of range.But the places most in need of telesurgery may also be the hardest places to deploy it. For starters, they may not have the high-quality network connections that remote surgery requires. Last June, experts from five major robotic corporations coauthored a list of technical standards for what they called “surgical-grade networks.” Mike Teodorescu, an information systems researcher with the University of Washington, says experts usually consider 100 milliseconds a safe threshold, but his team recently reported that many rural communities—and even most countries—struggle to meet it. Neuroradiologist Hal Rice of the Australian Stroke Alliance remotely operates Remedy’s N1 System.Remedy RoboticsAnd connectivity is only part of the problem. The smallest and most remote hospitals may lack other support needed for thrombectomy, including neurologists, catheterization teams, CT scanners, and specialized intensive care units. “It’s an infrastructure problem,” Teodorescu says.Medium-size regional hospitals may be a better target. Although they rarely offer thrombectomies, many already have the right support, experience, and patient volume. “There is a window for a thrombectomy center in a reasonably sized town where there is a stable economy,” says neurologist Bhatt.“Every minute wasted or delayed, 2 million brain cells are going to die.”–neurologist Archit BhattTelesurgery could do more than bring a distant specialist into the operating room. Jasdeep Bahra, an emergency medicine physician and lecturer in Kilifi, Kenya, worries that telerobotics could concentrate expertise in cities—widening the skill gap between rich cities and poor towns. The thrombectomy access problem is real, Bahra says, but lower income communities might benefit more if a system like Remedy’s could both treat patients and help train clinicians on-site. Sganga says Bahra’s idea fits Remedy’s mission.“Digital operating rooms where you have surgeons that are over your shoulder as a mentor,” Bahra says. “If that’s what we’re talking about in telesurgery, that is the future for me.” Measures of SuccessIf smoothness and precision aren’t clinical outcomes, as Bell told me, what would count as success? One target is safety. In more than 1 percent of thrombectomy operations, surgeons unintentionally slice open a blood vessel, a feared and sometimes fatal occurrence that Bell believes robots can prevent.In September 2025, Bell flew to Toronto for Remedy’s first test in human patients. Ten patients had consented to nonemergency procedures with Remedy’s robot. A neurosurgeon with a laptop would operate on six patients in an adjoining room; two elsewhere in the hospital; and two at an entirely different hospital across town. The Remedy team had experimented for years on silicone models, like the one I test-drove. Tests on lab animals suggested that, compared with a human, their robot collided less often with blood vessels and finished procedures faster. Neurosurgeon Vitor Mendes Pereira pilots the N1 system in 2024 for its first in-person robotic procedure, in Toronto. Remedy RoboticsIn the run-up to the Toronto procedures, the team measured internet latencies meticulously. Now they could only hope that their results would translate to human patients. “You’re never sure of how that leap is going to look, and it’s actually really hard to test,” Bell says. The night before the first operation, Bell barely slept.The procedures—recording images called angiograms from within patients’ blood vessels—were a success, and were completed nearly twice as quickly as the current standard. “I cannot explain the relief,” Bell says. Remedy considers speed one measure of success because it can allow doctors to treat more people. In 2025, neurosurgeon Vitor Mendes Pereira and his team at Unity Health Toronto perform the first fully remote robotic neurointerventional procedures. Remedy RoboticsA few months after my visit, Remedy announced US $62 million in new funding to support upcoming clinical trials. A portion of that investment came from the University of California system, which will host the clinical studies within a year, according to Bell. Remedy’s first trials will be on-site elective procedures, such as treating unruptured aneurysms, before expanding to remote surgeries in Australia or within the University of California system, near the end of 2027.Sganga emphasizes that the on-site trials are not just a stepping stone. He’s come to believe that Remedy’s success depends on improving operations even when the surgeon is in the room. “The remote solution—it’s very flashy, it’s very obvious what that patient need is—but in order to get there, we believe you have to solve a meaningful problem on-premise,” he says. “Otherwise, doctors simply won’t use the system, and if they don’t use the system, they’re not going to gain the expertise required to operate remotely.”The steerable catheter and computer-vision algorithm are two features he claims will improve thrombectomies, remote or not. And as trials ramp up, Sganga says they will look at factors beyond safety and surgical speed. “How quickly are doctors picking up on this new technique? How smoothly are the bedside interactions going? How quickly are folks able to set up the draping and interact with the tools?” he says. Bell has wondered whether to add a photo or video of the patient’s face to the surgeon’s display to preserve the emotional stakes of surgery.These ambitions draw Bell and Sganga’s 25-person company into uncharted waters. Beyond one-time demonstrations, no robotics company has earned full regulatory approval for telesurgery, and no patient has been operated on remotely in the United States. Remedy hopes to change that. Decades after stroke research conquered the “therapeutic nihilism” that once defined the field, Remedy hopes telesurgery can bring thrombectomy to everyone who needs it, no matter where they might be. Read More